Explained in 4 Charts: What By-Elections Tell Us & What They Don’t

60% of bypolls since 2019 were won by the party in power in the state-level.

51 Assembly seats and 2 Lok Sabha constituencies will vote in the by-elections on 21 October. 

(Photo: The Quint)

As many as 51 Assembly constituencies and two Lok Sabha constituencies, spread across 18 states, will be voting in the crucial by-elections on 21 October, along with the Assembly polls in Maharashtra and in Haryana. Here’s how the 51 Assembly seats are distributed:

• Arunachal Pradesh: 1
• Assam: 4
• Bihar: 5
• Chhattisgarh: 1
• Gujarat: 6
• Himachal Pradesh: 2
• Kerala: 5
• Madhya Pradesh: 1
• Meghalaya: 1
• Odisha: 1
• Puducherry: 1
• Punjab: 4
• Rajasthan: 2
• Sikkim: 3
• Tamil Nadu: 2
• Telangana: 1
• Uttar Pradesh: 11

In addition to these, there will also be bypolls in two Lok Sabha seats: Samastipur in Bihar and Satara in Maharashtra.

As these 53 seats – 51 Assembly and 2 Lok Sabha constituencies – are distributed across 18 states, many are calling this a mini-national election, the result of which could reflect the public mood over key national issues like the abrogation of Article 370 or the economic slowdown.

But do bypolls really reflect the national sentiment or are they fought on the basis of local factors?

Here are a few inferences we can draw by analysing the previous ten years’ by-election results.

1. Union Govt’s Popularity Not a Big Factor

We have collected data from all the bypolls held between 2009 and 2019. Over 450 bypolls were held during this period, including the UPA’s second tenure (2009-2014), the Narendra Modi-led government’s first tenure as well as the first few months of his second term (2014-till date).

The NDA has won close to 40 percent and lost 60 percent bypolls held since it came to power in 2014. This success rate is about five percentage points more than that of the UPA’s 34.4 percent between 2009 and 2014.

Now, we all know that the UPA was decimated in the 2014 Lok Sabha elections while the NDA won a huge majority in 2019. So, this relatively small gap in their bypoll success rate on the one hand and the huge contrast in their political fortunes on the other, indicates that by-election results don’t really give us an idea of the popularity of the Central government. And, there is no major difference between Lok Sabha and Assembly bypolls in this respect.

2. Incumbents in States Are Advantaged

More than to the Centre, by-elections – both Assembly and Lok Sabha – seem more connected to state-level factors. Past results indicate that the party in power in the state has an advantage during bypolls. For instance, the recent bypolls in Pala (Kerala), Badharghat (Tripura), Dantewada (Chhattisgarh) and Hamirpur (Uttar Pradesh) were won by the party in power in the respective states.

Out of over 450 bypolls that have been held from 2009 till date, 272 or nearly 60 percent were won by the party in power at the state level.

Psephologist and founder of survey agency CVoter, Yashwant Deshmukh also holds the same view.

“Parties in power in the state are at an advantage (in a bypoll) simply because the entire state machinery gets focused on that particular area, including intense campaign by the entire star cast which in a general election gets spread too thin,” he told The Quint.

However, Deshmukh adds a caveat which we shall discuss a little later.

3. There Are State-wise Variations

The advantage for state governments isn’t uniform across India. There are some variables which make a difference:

1. State-wise variations
2. Popularity of the state government

Let’s look at the first aspect in this section.

Data from all bypolls between 2009 and 2019 reveal that in six states, all the bypolls were won by the government in power at the state level: Arunachal Pradesh, Telangana, Nagaland, Mizoram, Odisha and Goa.

In seven more states, the success rate of incumbent state governments was 75 percent or more: Punjab (92 percent), Uttarakhand (89 percent), Tripura (83 percent), Chhattisgarh (83 percent), Assam (79 percent), Gujarat (77 percent), Kerala (75 percent).

This success rate is irrespective of whether the party in power at the state level is the BJP, Congress or a regional party.

States like Odisha, Gujarat, and Telangana have a history of re-electing governments so the high success rate in these states could be due to the popularity or entrenchment of the ruling party and weakness of the Opposition.

However, in some other states, it could just be the advantage of having the state government on one’s side and the political culture of that state, which could involve a closer coordination between the government machinery and the party. For instance, the Left Democratic Front ended up winning the Pala seat last month despite having never won the seat in five decades. The Left performed poorly in Pala even in the 2019 Lok Sabha polls, so the bypoll win couldn’t have been due to some sudden surge in popularity and rather, can be attributed to the advantage of being in power.

Punjab’s case is similar. The Shiromani Akali Dal-BJP alliance won two out of three bypolls between 2014 and 2017 and yet, got defeated in the 2017 Assembly polls, indicating that their by-election wins had little to do with their popularity on ground.

On the other hand, in Jharkhand where the incumbent party – be it BJP, Congress or JMM – has ended up losing 75 percent of the bypolls in the last 10 years

4. Popularity of State Government a Major Factor

However, there are cases where bypolls indicate the complete lack of popularity of a state government and even being in power doesn’t lead to victory.

This brings us to Yashwant Deshmukh’s caveat that we mentioned earlier.

“State governments’ advantage in bypolls is applicable only when the it is still having a half-decent image or in honeymoon period. Else the bypolls are known to give the most stinging anti-incumbent verdict even in the face of entire combined might of state machinery and campaign star-cast,” he told The Quint.

This is clear in the manner in which the Left Front lost nearly 90 percent of the bypolls in West Bengal between 2009 and 2011, when it was eventually voted out.

Another case in point is Andhra Pradesh between 2009 and 2014. Despite being in power, the Congress lost 93 percent of the bypolls in the state in this period, which provided a clear indication of the party’s complete collapse in the state.

Bypolls also indicated the AIADMK’s decline in Tamil Nadu after the death of J Jayalalithaa in December 2016. Between 2011 and 2016, the ruling AIADMK won each and every bypoll in the state. But since 2016, it has lost around 60 percent of the by-elections there. This is much lower than even the DMK’s last two years in power (2009-11) when it won over 80 percent of the bypolls.

If nothing much changes in Tamil Nadu in the next couple of years, the bypolls indicate that the AIADMK could be staring at a massive defeat in the next Assembly elections.

In contrast to these three states is Uttar Pradesh, where bypolls don’t seem to reveal any major patterns. The BSP’s success rate between 2009 and 2012 and the SP’s from 2012 to 2017 is around the national average of 60 percent. On the other hand, the BJP has fared poorly in bypolls, winning just 40 percent since it came to power.

It seems that in larger states like Uttar Pradesh and Maharashtra, with relatively robust local party competition, bypolls to a great extent are decided by constituency-specific dynamics and not so much the performance of the state government.

What Can Happen in the 53 Upcoming Bypolls?

• With the bypolls in Karnataka being postponed to December, the largest share of seats where by-elections will be held on 21 October is Uttar Pradesh with 11 seats. If the past results are any indication, these seats could end up giving a mixed verdict based on local alignments. If that doesn’t happen and there is a decisive shift in favour of or against the BJP, it could mean that there is a larger state-level sentiment in that direction.
• The same pattern may apply to the five Assembly and one Lok Sabha seat in Bihar.
• Majority among the six seats in Gujarat, five in Kerala, four each in Punjab and Assam and one each in Odisha, Telangana, Arunachal Pradesh and Chhattisgarh could go in favour of the incumbent party in the respective states, if past results are anything to go by. If state incumbent parties don’t win a majority of these seats, it could be an indication of an anti-incumbency sentiment strong enough to neutralise their advantage.
• The two bypolls in Tamil Nadu could be in line with the broader decline of the AIADMK in the state.
• Nationally, the bypolls are unlikely to reflect a major trend either for or against the Narendra Modi-led government.